You’re out of free articles.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Sign In or Create an Account.
By continuing, you agree to the Terms of Service and acknowledge our Privacy Policy
Welcome to Heatmap
Thank you for registering with Heatmap. Climate change is one of the greatest challenges of our lives, a force reshaping our economy, our politics, and our culture. We hope to be your trusted, friendly, and insightful guide to that transformation. Please enjoy your free articles. You can check your profile here .
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Subscribe to get unlimited Access
Hey, you are out of free articles but you are only a few clicks away from full access. Subscribe below and take advantage of our introductory offer.
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Create Your Account
Please Enter Your Password
Forgot your password?
Please enter the email address you use for your account so we can send you a link to reset your password:
AI has already changed weather forecasting forever.
It’s been a wild few years in the typically tedious world of weather predictions. For decades, forecasts have been improving at a slow and steady pace — the standard metric is that every decade of development leads to a one-day improvement in lead time. So today, our four-day forecasts are about as accurate as a one-day forecast was 30 years ago. Whoop-de-do.
Now thanks to advances in (you guessed it) artificial intelligence, things are moving much more rapidly. AI-based weather models from tech giants such as Google DeepMind, Huawei, and Nvidia are now consistently beating the standard physics-based models for the first time. And it’s not just the big names getting into the game — earlier this year, the 27-person team at Palo Alto-based startup Windborne one-upped DeepMind to become the world’s most accurate weather forecaster.
“What we’ve seen for some metrics is just the deployment of an AI-based emulator can gain us a day in lead time relative to traditional models,” Daryl Kleist, who works on weather model development at the National Oceanic and Atmospheric Administration, told me. That is, today’s two-day forecast could be as accurate as last year’s one-day forecast.
All weather models start by taking in data about current weather conditions. But from there, how they make predictions varies wildly. Traditional weather models like the ones NOAA and the European Centre for Medium-Range Weather Forecasts use rely on complex atmospheric equations based on the laws of physics to predict future weather patterns. AI models, on the other hand, are trained on decades of prior weather data, using the past to predict what will come next.
Kleist told me he certainly saw AI-based weather forecasting coming, but the speed at which it’s arriving and the degree to which these models are improving has been head-spinning. “There's papers coming out in preprints almost on a bi-weekly basis. And the amount of skill they've been able to gain by fine tuning these things and taking it a step further has been shocking, frankly,” he told me.
So what changed? As the world has seen with the advent of large language models like ChatGPT, AI architecture has gotten much more powerful, period. The weather models themselves are also in a cycle of continuous improvement — as more open source weather data becomes available, models can be retrained. Plus, the cost of computing power has come way down, making it possible for a small company like Windborne to train its industry-leading model.
Founded by a team of Stanford students and graduates in 2019, Windborne used off-the-shelf Nvidia gaming GPUs to train its AI model, called WeatherMesh — something the company’s CEO and co-founder, John Dean, told me wouldn’t have been possible five years ago. The company also operates its own fleet of advanced weather balloons, which gather data from traditionally difficult-to-access areas.
Standard weather balloons without onboard navigation typically ascend too high, overinflate, and pop within a matter of hours (thus becoming environmental waste, sad!). Since it’s expensive to do launches at sea or in areas without much infrastructure, there’s vast expanses of the globe where most balloons aren’t gathering any data at all.
Satellites can help, of course. But because they’re so far away, they can’t provide the same degree of fidelity. With modern electronics, though, Windborne found it could create a balloon that autonomously changes altitude and navigates to its intended target by venting gas to descend and dropping ballast to ascend.
“We basically took a lot of the innovations that lead to smartphones, global satellite communications, all of the last 20 years of progress in consumer electronics and other things and applied that to balloons,” Dean told me. In the past, the electronics needed to control Windborne’s system would have been too heavy — the balloon wouldn’t have gotten off the ground. But with today’s tiny tech, they can stay aloft for up to 40 days. Eventually, the company aims to recover and reuse at least 80% of its balloons.
The longer airtime allows Windborne to do more with less. While globally there are more than 1,000 conventional weather balloons launched every day, Dean told me, “We collect roughly on the order of 10% or 20% of the data that NOAA collects every day with only 100 launches per month.” In fact, NOAA is a customer of the startup — Windborne already makes millions in revenue selling its weather balloon data to various government agencies.
Now, with a potentially historic hurricane season ramping up, Windborne has the potential to provide the most accurate data on when and where a storm will touch down.
Earlier this year, the company used WeatherMesh to run a case study on Hurricane Ian, the Category 5 storm that hit Florida in September 2022, leading to over 150 fatalities and $112 billion in damages. Using only weather data that was publicly available at the time, the company looked at how accurately its model (had it existed back then) would have tracked the hurricane.
Very accurately, it turns out. Windborne’s predictions aligned neatly with the storm’s actual path, while the National Weather Service’s model was off by hundreds of kilometers. That impressed Khosla Ventures, which led the company’s $15 million Series A funding round earlier this month. “We haven’t seen meaningful innovation in weather since The Weather Channel in the 90s. Yet it’s a $100 billion market that touches essentially every industry,” Sven Strohband, a partner and managing director at Khosla Ventures, told me via email.
With this new funding, Windborne is scaling up its fleet of balloons as it prepares to commercialize. The money will also help Windborne advance its forecasting model, though Dean told me robust data collection is ultimately what will set the company apart. “In any kind of AI industry, whoever has the top benchmark at any given time, it’s going to fluctuate,” Dean said. “What matters is the model plus the unique datasets.”
Unlike Windborne, the tech giants with AI-based weather models — including, most recently, Microsoft — aren’t gathering their own data, instead drawing solely on publicly accessible information from legacy weather agencies.
But these agencies are starting to get into the game, too. The European Centre for Medium-Range Weather Forecasts has already created its own AI-based model, the Artificial Intelligence/Integrated Forecasting System, which it runs in parallel to its traditional model. NOAA, while a bit behind, is also looking to follow suit.
“In the end, we know we can't rely on these big tech companies to just keep developing stuff in good faith to give to us for free,” Kleist told me. Right now, many of the top AI-based weather models are open source. But who knows if that will last? “It's our mission to save lives and property. And we have to figure out how to do some of this development and operationalize it from our side, ourselves,” Kleist said, explaining that NOAA is currently prototyping some of its own AI-based models.
All of these agencies are in the early stages of AI modeling, which is why you likely haven’t noticed weather predictions making a pronounced leap in accuracy as of late. It’s all still considered quite experimental. “Physical models, the pro is we know the underlying assumptions we make. We understand them. We have decades of history of developing them and using them in operational settings,” Kleist told me. AI-based models are much more of a black box, and there’s questions surrounding how well they will perform when it comes to predicting rare weather events, for which there might be little to no historical data for the model to reference.
That hesitation might not last long, though. “To me it’s fairly obvious that most of the forecasts that would actually be used by users in the future will come from machine learning models,” Peter Dueben, head of Earth systems modeling at the European Centre for Medium Range Weather Forecasting, told me. “If you just want to get the weather forecast for the temperature in California tomorrow, then the machine learning model is typically the better choice,” he added.
That increased accuracy is going to matter a lot, not just for the average weather watcher, but also for specific industries and interest groups for whom precise predictions are paramount. “We can tailor the actual models to particular sectors, whether it's agriculture, energy, transportation,” Kleist told me, “and come up with information that's going to be at a very granular, specific level to a particular interest.” Think grid operators or renewable power generators who need to forecast demand or farmers trying to figure out the best time to irrigate their fields or harvest crops.
A major (and perhaps surprising) reason this type of customization is so easy is because once AI-based weather models are trained, they’re actually orders of magnitude cheaper and less computationally intensive to run than traditional models. All of this means, Kleist told me, that AI-based weather models are “going to be fundamentally foundational for what we do in the future, and will open up avenues to things we couldn't have imagined using our current physical-based modeling.”
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
New research out today shows a 10-fold increase in smoke mortality related to climate change from the 1960s to the 2010.
If you are one of the more than 2 billion people on Earth who have inhaled wildfire smoke, then you know firsthand that it is nasty stuff. It makes your eyes sting and your throat sore and raw; breathe in smoke for long enough, and you might get a headache or start to wheeze. Maybe you’ll have an asthma attack and end up in the emergency room. Or maybe, in the days or weeks afterward, you’ll suffer from a stroke or heart attack that you wouldn’t have had otherwise.
Researchers are increasingly convinced that the tiny, inhalable particulate matter in wildfire smoke, known as PM2.5, contributes to thousands of excess deaths annually in the United States alone. But is it fair to link those deaths directly to climate change?
A new study published Monday in Nature Climate Change suggests that for a growing number of cases, the answer should be yes. Chae Yeon Park, a climate risk modeling researcher at Japan’s National Institute for Environmental Studies, looked with her colleagues at three fire-vegetation models to understand how hazardous emissions changed from 1960 to 2019, compared to a hypothetical control model that excluded historical climate change data. They found that while fewer than 669 deaths in the 1960s could be attributed to climate change globally, that number ballooned to 12,566 in the 2010s — roughly a 20-fold increase. The proportion of all global PM2.5 deaths attributable to climate change jumped 10-fold over the same period, from 1.2% in the 1960s to 12.8% in the 2010s.
“It’s a timely and meaningful study that informs the public and the government about the dangers of wildfire smoke and how climate change is contributing to that,” Yiqun Ma, who researches the intersection of climate change, air pollution, and human health at the Yale School of Medicine, and who was not involved in the Nature study, told me.
The study found the highest climate change-attributable fire mortality values in South America, Australia, and Europe, where increases in heat and decreases in humidity were also the greatest. In the southern hemisphere of South America, for example, the authors wrote that fire mortalities attributable to climate change increased from a model average of 35% to 71% between the 1960s and 2010s, “coinciding with decreased relative humidity,” which dries out fire fuels. For the same reason, an increase in relative humidity lowered fire mortality in other regions, such as South Asia. North America exhibited a less dramatic leap in climate-related smoke mortalities, with climate change’s contribution around 3.6% in the 1960s, “with a notable rise in the 2010s” to 18.8%, Park told me in an email.
While that’s alarming all on its own, Ma told me there was a possibility that Park’s findings might actually be too conservative. “They assume PM2.5 from wildfire sources and from other sources” — like from cars or power plants — “have the same toxicity,” she explained. “But in fact, in recent studies, people have found PM2.5 from fire sources can be more toxic than those from an urban background.” Another reason Ma suspected the study’s numbers might be an underestimate was because the researchers focused on only six diseases that have known links to PM2.5 exposure: chronic obstructive pulmonary disease, lung cancer, coronary heart disease, type 2 diabetes, stroke, and lower respiratory infection. “According to our previous findings [at the Yale School of Medicine], other diseases can also be influenced by wildfire smoke, such as mental disorders, depression, and anxiety, and they did not consider that part,” she told me.
Minghao Qiu, an assistant professor at Stony Brook University and one of the country’s leading researchers on wildfire smoke exposure and climate change, generally agreed with Park’s findings, but cautioned that there is “a lot of uncertainty in the underlying numbers” in part because, intrinsically, wildfire smoke exposure is such a complicated thing to try to put firm numbers to. “It’s so difficult to model how climate influences wildfire because wildfire is such an idiosyncratic process and it’s so random, ” he told me, adding, “In general, models are not great in terms of capturing wildfire.”
Despite their few reservations, both Qiu and Ma emphasized the importance of studies like Park’s. “There are no really good solutions” to reduce wildfire PM2.5 exposure. You can’t just “put a filter on a stack” as you (sort of) can with power plant emissions, Qiu pointed out.
Even prescribed fires, often touted as an important wildfire mitigation technique, still produce smoke. Park’s team acknowledged that a whole suite of options would be needed to minimize future wildfire deaths, ranging from fire-resilient forest and urban planning to PM2.5 treatment advances in hospitals. And, of course, there is addressing the root cause of the increased mortality to begin with: our warming climate.
“To respond to these long-term changes,” Park told me, “it is crucial to gradually modify our system.”
On the COP16 biodiversity summit, Big Oil’s big plan, and sea level rise
Current conditions: Record rainfall triggered flooding in Roswell, New Mexico, that killed at least two people • Storm Ashley unleashed 80 mph winds across parts of the U.K. • A wildfire that broke out near Oakland, California, on Friday is now 85% contained.
Forecasters hadn’t expected Hurricane Oscar to develop into a hurricane at all, let alone in just 12 hours. But it did. The Category 1 storm made landfall in Cuba on Sunday, hours after passing over the Bahamas, bringing intense rain and strong winds. Up to a foot of rainfall was expected. Oscar struck while Cuba was struggling to recover from a large blackout that has left millions without power for four days. A second system, Tropical Storm Nadine, made landfall in Belize on Saturday with 60 mph winds and then quickly weakened. Both Oscar and Nadine developed in the Atlantic on the same day.
Hurricane OscarAccuWeather
The COP16 biodiversity summit starts today in Cali, Colombia. Diplomats from 190 countries will try to come up with a plan to halt global biodiversity loss, aiming to protect 30% of land and sea areas and restore 30% of degraded ecosystems by 2030. Discussions will revolve around how to monitor nature degradation, hold countries accountable for their protection pledges, and pay for biodiversity efforts. There will also be a big push to get many more countries to publish national biodiversity strategies. “This COP is a test of how serious countries are about upholding their international commitments to stop the rapid loss of biodiversity,” said Crystal Davis, Global Director of Food, Land, and Water at the World Resources Institute. “The world has no shot at doing so without richer countries providing more financial support to developing countries — which contain most of the world’s biodiversity.”
A prominent group of oil and gas producers has developed a plan to roll back environmental rules put in place by President Biden, The Washington Post reported. The paper got its hands on confidential documents from the American Exploration and Production Council (AXPC), which represents some 30 producers. The documents include draft executive orders promoting fossil fuel production for a newly-elected President Trump to sign if he takes the White House in November, as well as a roadmap for dismantling many policies aimed at getting oil and gas producers to disclose and curb emissions. AXPC’s members, including ExxonMobil, ConocoPhillips, and Hess, account for about half of the oil and gas produced in the U.S., the Post reported.
A new report from the energy think tank Ember looks at how the uptake of electric vehicles and heat pumps in the U.K. is affecting oil and gas consumption. It found that last year the country had 1.5 million EVs on the road, and 430,000 residential heat pumps in homes, and the reduction in fossil fuel use due to the growth of these technologies was equivalent to 14 million barrels of oil, or about what the U.K. imports over a two-week span. This reduction effect will be even stronger as more and more EVs and heat pumps are powered by clean energy. The report also found that even though power demand is expected to rise, efficiency gains from electrification and decarbonization will make up for this, leading to an overall decline in energy use and fossil fuel consumption.
Ember
The world’s sea levels are projected to rise by more than 6 inches on average over the next 30 years if current trends continue, according to a new study published in the journal Nature. “Such rates would represent an evolving challenge for adaptation efforts,” the authors wrote. By examining satellite data, the researchers found that sea levels have risen by about .4 inches since 1993, and that they’re rising faster now than they were then. In 1993 the seas were rising by about .08 inches per year, and last year they were rising at .17 inches per year. These are averages, of course, and some areas are seeing much more extreme changes. For example, areas around Miami, Florida, have already seen sea levels rise by 6 inches over the last 31 years.
“As the climate crisis grows more urgent, restoring faith in government will be more important than ever.” –Paul Waldman writing for Heatmap about the profound implications of America becoming a low-trust society.
That means big, bad things for disaster relief — and for climate policy in general.
When Hurricanes Helene and Milton swept through the Southeast, small-government conservatives demanded fast and effective government service, in the form of relief operations organized by the Federal Emergency Management Agency. Yet even as the agency was scrambling to meet the need, it found itself targeted by far-right militias, who prevented it from doing its job because they had been led by cynical politicians to believe it wasn't doing its job.
It’s almost a law of nature, or at least of politics, that when government does its job, few people notice — only when it screws up does everyone pay attention. While this is nothing new in itself, it has increasingly profound implications for the future of government-driven climate action. While that action comes in many forms and can be sold to the public in many ways, it depends on people having faith that when government steps in — whether to create new regulations, invest in new technologies, or provide benefits for climate-friendly choices — it knows what it’s doing and can accomplish its goals.
As the climate crisis grows more urgent, restoring faith in government will be more important than ever. Unfortunately, simply doing the right things — like responding competently to disasters — won’t be enough to convince people that the next climate initiative will do what it’s supposed to.
The number of people expressing faith in government today is nearly as low as it has been in the half-century pollsters have been asking the question. That trust has bounced up and down a bit — it rose after September 11, then fell again during the disastrous Iraq War — but for the last decade and half, only around 20% of Americans say they trust the government most of the time.
It’s partisan, of course: People express more trust when their party controls the White House. And the decline of trust reaches beyond the government. Faith in most of the key institutions of American life — business, education, religion, news media — has fallen in recent decades, sometimes for good reason. The net result is a public skeptical that those in authority have the ability to solve complex problems.
Changing that perspective is extraordinarily difficult, often because of the nature of good and bad news: The former usually happens slowly and invisibly, while the latter often happens dramatically and all at once.
Take the program created in the Energy Department under George W. Bush to provide loans to innovative energy technologies. If most Americans had heard of it, it was because of one company: Solyndra, a manufacturer of innovative but overly expensive solar panels. Undercut by a decline in prices of traditional panels, the company went under, and its $535 million loan was never repaid. Republicans made Solyndra’s failure into a major controversy, claiming that the program showed that government investment in green technology was corrupt, ineffective, and wasteful.
What few people heard about was that the loan program overall not only turned a profit at the time (and for what it’s worth, it still does), it also provided help to many successful companies, even if a few failed — as any venture capital investor could tell you is inevitable. The successes included Tesla, which used its federal loan to ramp up production of the sedans that would turn it from a niche manufacturer of electric roadsters into what it is today. Needless to say, Elon Musk does not advertise the fact that his success was built on government help.
More recently, the hurricane response has shown how partisan polarization can be used to undermine trust in government — especially when Donald Trump is involved. Trump took the opportunity of the hurricanes to accuse the federal government of being both political and partisan, delivering help only to those areas that vote for Democrats. Soon after, he promised to do precisely what he falsely accused the Biden administration of doing, saying that if he is president again, he will withhold disaster aid from California unless Gov. Gavin Newsom changes the state’s water policies to be more to Trump’s liking. “And we’ll say, Gavin, if you don’t do it, we’re not giving any of that fire money that we send you all the time for all the fire, forest fires that you have,” Trump said. And in fact, in his first term Trump did try to withhold disaster aid from blue states.
What sounds like hypocrisy is actually something much more pernicious. As he often does, Trump is arguing not that he is clean and his opponents are dirty, but that everyone is dirty, and it’s just a question of whether government is in the hands of our team or their team. When he says he’ll “drain the swamp,” he’s telling people both that government is corrupt, and the answer is merely to change who gets the spoils. If you believe him, you’ll have no trust in government whatsoever, even if you might think he’ll use it in a way you’ll approve of.
We’ve seen again and again that people want government to perform well and get angry when it doesn’t, but they don’t reward competence when it happens. Which is why making sure systems operate properly and problems are solved is necessary but not sufficient to win back trust. Government’s advocates — especially those who are counting on it to undertake ambitious climate action both now and in the future — need not only to deliver, they have to get better at, for lack of a better word, propaganda. Policy success is not its own advertisement. And despite his ample policy achievements, Joe Biden has not been a charismatic and effective messenger — on the role of government, or much else.
Ronald Reagan used to say that the most frightening words in the English language were “I’m from the government and I’m here to help”; the oft-repeated quip was at the center of his incredibly successful effort to delegitimize government in the eyes of voters. To reverse the decline of trust so people will believe that government has the knowledge and ability to tackle climate change, the public needs to be reminded — often and repeatedly — of what government does well.
Touting past and present successes on climate — and disaster relief, and so many other ways the government solves problems every day — is essential to building support for future climate initiatives. Those successes are all around, it’s just that most people never hear about them or take them for granted. But promoting government as an engine of positive change should be as high a priority for climate advocates, including those who hold public office, as discrediting government was for Reagan and is for Trump.